Predictive Processing During Cue-Outcome Associative Learning in Autistic Children.

Area of Science:

• Neuroscience
• Developmental Psychology
• Cognitive Science

Background:

• Predictive coding theories suggest autism involves over-adjustment to prediction errors, impacting belief updating.
• This atypical processing may hinder stable world model construction but could offer benefits.

Purpose of the Study:

• Investigate if autistic children update learning faster towards subtle cues after unexpected events.
• Explore the link between language proficiency and predictive performance in autistic and non-autistic children.

Main Methods:

• Recorded anticipatory fixations and explicit predictions in three associative learning tasks.
• Utilized deterministic and probabilistic contingencies, including a task where a subtle cue was overshadowed by a salient one.

Main Results:

• Both groups learned using the salient cue, with no evidence of updating towards the subtle cue.
• Autistic children made less accurate explicit predictions than non-autistic children.
• Language proficiency correlated with prediction accuracy in non-autistic children, but not in autistic children.

Conclusions:

• Results do not support over-adjustment to prediction errors in autistic children.
• Controlling for general performance in associative learning is crucial for predictive processing studies.
• Further research is needed on predictive processing and language development in autism.